A summary of methods yielding information about the generation and configuration of the geomagnetic main field is presented with special focus on complications concerning these methods. A global source model constructed with the help of machine learning (and deep learning) is proposed to mitigate these issues, in particular the uncertainties caused by vigorous convection and small scale fields.

中文翻译：

---

识别3D外核流的地磁特征的最新进展

总结了产生有关地磁主场的生成和配置信息的方法的摘要，并特别关注与这些方法有关的复杂性。为了缓解这些问题，特别是在机器学习（和深度学习）的帮助下，提出了一种全局源模型，以缓解这些问题，特别是由于剧烈对流和小规模场造成的不确定性。